2023-10-10 22:00:43,304 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,306 Model: "SequenceTagger( (embeddings): ByT5Embeddings( (model): T5EncoderModel( (shared): Embedding(384, 1472) (encoder): T5Stack( (embed_tokens): Embedding(384, 1472) (block): ModuleList( (0): T5Block( (layer): ModuleList( (0): T5LayerSelfAttention( (SelfAttention): T5Attention( (q): Linear(in_features=1472, out_features=384, bias=False) (k): Linear(in_features=1472, out_features=384, bias=False) (v): Linear(in_features=1472, out_features=384, bias=False) (o): Linear(in_features=384, out_features=1472, bias=False) (relative_attention_bias): Embedding(32, 6) ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (1): T5LayerFF( (DenseReluDense): T5DenseGatedActDense( (wi_0): Linear(in_features=1472, out_features=3584, bias=False) (wi_1): Linear(in_features=1472, out_features=3584, bias=False) (wo): Linear(in_features=3584, out_features=1472, bias=False) (dropout): Dropout(p=0.1, inplace=False) (act): NewGELUActivation() ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) (1-11): 11 x T5Block( (layer): ModuleList( (0): T5LayerSelfAttention( (SelfAttention): T5Attention( (q): Linear(in_features=1472, out_features=384, bias=False) (k): Linear(in_features=1472, out_features=384, bias=False) (v): Linear(in_features=1472, out_features=384, bias=False) (o): Linear(in_features=384, out_features=1472, bias=False) ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (1): T5LayerFF( (DenseReluDense): T5DenseGatedActDense( (wi_0): Linear(in_features=1472, out_features=3584, bias=False) (wi_1): Linear(in_features=1472, out_features=3584, bias=False) (wo): Linear(in_features=3584, out_features=1472, bias=False) (dropout): Dropout(p=0.1, inplace=False) (act): NewGELUActivation() ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) ) (final_layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) (locked_dropout): LockedDropout(p=0.5) (linear): Linear(in_features=1472, out_features=17, bias=True) (loss_function): CrossEntropyLoss() )" 2023-10-10 22:00:43,306 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,307 MultiCorpus: 1166 train + 165 dev + 415 test sentences - NER_HIPE_2022 Corpus: 1166 train + 165 dev + 415 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fi/with_doc_seperator 2023-10-10 22:00:43,307 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,307 Train: 1166 sentences 2023-10-10 22:00:43,307 (train_with_dev=False, train_with_test=False) 2023-10-10 22:00:43,307 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,307 Training Params: 2023-10-10 22:00:43,307 - learning_rate: "0.00016" 2023-10-10 22:00:43,307 - mini_batch_size: "8" 2023-10-10 22:00:43,307 - max_epochs: "10" 2023-10-10 22:00:43,307 - shuffle: "True" 2023-10-10 22:00:43,307 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,307 Plugins: 2023-10-10 22:00:43,307 - TensorboardLogger 2023-10-10 22:00:43,307 - LinearScheduler | warmup_fraction: '0.1' 2023-10-10 22:00:43,308 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,308 Final evaluation on model from best epoch (best-model.pt) 2023-10-10 22:00:43,308 - metric: "('micro avg', 'f1-score')" 2023-10-10 22:00:43,308 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,308 Computation: 2023-10-10 22:00:43,308 - compute on device: cuda:0 2023-10-10 22:00:43,308 - embedding storage: none 2023-10-10 22:00:43,308 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,308 Model training base path: "hmbench-newseye/fi-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-1" 2023-10-10 22:00:43,308 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,308 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:00:43,308 Logging anything other than scalars to TensorBoard is currently not supported. 2023-10-10 22:00:52,394 epoch 1 - iter 14/146 - loss 2.83088289 - time (sec): 9.08 - samples/sec: 530.64 - lr: 0.000014 - momentum: 0.000000 2023-10-10 22:01:00,821 epoch 1 - iter 28/146 - loss 2.82628929 - time (sec): 17.51 - samples/sec: 492.26 - lr: 0.000030 - momentum: 0.000000 2023-10-10 22:01:10,396 epoch 1 - iter 42/146 - loss 2.81614191 - time (sec): 27.09 - samples/sec: 505.57 - lr: 0.000045 - momentum: 0.000000 2023-10-10 22:01:19,597 epoch 1 - iter 56/146 - loss 2.80155779 - time (sec): 36.29 - samples/sec: 497.39 - lr: 0.000060 - momentum: 0.000000 2023-10-10 22:01:28,047 epoch 1 - iter 70/146 - loss 2.77881902 - time (sec): 44.74 - samples/sec: 485.62 - lr: 0.000076 - momentum: 0.000000 2023-10-10 22:01:36,553 epoch 1 - iter 84/146 - loss 2.73383396 - time (sec): 53.24 - samples/sec: 482.56 - lr: 0.000091 - momentum: 0.000000 2023-10-10 22:01:44,474 epoch 1 - iter 98/146 - loss 2.67503100 - time (sec): 61.16 - samples/sec: 478.46 - lr: 0.000106 - momentum: 0.000000 2023-10-10 22:01:54,877 epoch 1 - iter 112/146 - loss 2.57315899 - time (sec): 71.57 - samples/sec: 485.69 - lr: 0.000122 - momentum: 0.000000 2023-10-10 22:02:03,653 epoch 1 - iter 126/146 - loss 2.49591763 - time (sec): 80.34 - samples/sec: 482.87 - lr: 0.000137 - momentum: 0.000000 2023-10-10 22:02:12,893 epoch 1 - iter 140/146 - loss 2.41324023 - time (sec): 89.58 - samples/sec: 479.18 - lr: 0.000152 - momentum: 0.000000 2023-10-10 22:02:16,311 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:02:16,311 EPOCH 1 done: loss 2.3834 - lr: 0.000152 2023-10-10 22:02:21,884 DEV : loss 1.2905865907669067 - f1-score (micro avg) 0.0 2023-10-10 22:02:21,893 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:02:31,067 epoch 2 - iter 14/146 - loss 1.32629947 - time (sec): 9.17 - samples/sec: 499.38 - lr: 0.000158 - momentum: 0.000000 2023-10-10 22:02:40,124 epoch 2 - iter 28/146 - loss 1.29112932 - time (sec): 18.23 - samples/sec: 488.68 - lr: 0.000157 - momentum: 0.000000 2023-10-10 22:02:49,458 epoch 2 - iter 42/146 - loss 1.15322127 - time (sec): 27.56 - samples/sec: 476.15 - lr: 0.000155 - momentum: 0.000000 2023-10-10 22:02:59,632 epoch 2 - iter 56/146 - loss 1.05695895 - time (sec): 37.74 - samples/sec: 467.96 - lr: 0.000153 - momentum: 0.000000 2023-10-10 22:03:10,417 epoch 2 - iter 70/146 - loss 0.98742812 - time (sec): 48.52 - samples/sec: 472.43 - lr: 0.000152 - momentum: 0.000000 2023-10-10 22:03:19,883 epoch 2 - iter 84/146 - loss 0.93247747 - time (sec): 57.99 - samples/sec: 463.34 - lr: 0.000150 - momentum: 0.000000 2023-10-10 22:03:29,139 epoch 2 - iter 98/146 - loss 0.88912171 - time (sec): 67.24 - samples/sec: 457.24 - lr: 0.000148 - momentum: 0.000000 2023-10-10 22:03:38,219 epoch 2 - iter 112/146 - loss 0.85474722 - time (sec): 76.32 - samples/sec: 452.98 - lr: 0.000147 - momentum: 0.000000 2023-10-10 22:03:46,844 epoch 2 - iter 126/146 - loss 0.82173264 - time (sec): 84.95 - samples/sec: 456.28 - lr: 0.000145 - momentum: 0.000000 2023-10-10 22:03:55,388 epoch 2 - iter 140/146 - loss 0.79360195 - time (sec): 93.49 - samples/sec: 459.36 - lr: 0.000143 - momentum: 0.000000 2023-10-10 22:03:58,923 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:03:58,923 EPOCH 2 done: loss 0.7905 - lr: 0.000143 2023-10-10 22:04:04,751 DEV : loss 0.3894149363040924 - f1-score (micro avg) 0.0 2023-10-10 22:04:04,760 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:04:13,057 epoch 3 - iter 14/146 - loss 0.51507142 - time (sec): 8.29 - samples/sec: 425.69 - lr: 0.000141 - momentum: 0.000000 2023-10-10 22:04:22,691 epoch 3 - iter 28/146 - loss 0.41692054 - time (sec): 17.93 - samples/sec: 477.51 - lr: 0.000139 - momentum: 0.000000 2023-10-10 22:04:31,502 epoch 3 - iter 42/146 - loss 0.43730739 - time (sec): 26.74 - samples/sec: 477.83 - lr: 0.000137 - momentum: 0.000000 2023-10-10 22:04:40,274 epoch 3 - iter 56/146 - loss 0.42787094 - time (sec): 35.51 - samples/sec: 470.86 - lr: 0.000136 - momentum: 0.000000 2023-10-10 22:04:49,689 epoch 3 - iter 70/146 - loss 0.41822906 - time (sec): 44.93 - samples/sec: 467.47 - lr: 0.000134 - momentum: 0.000000 2023-10-10 22:04:58,317 epoch 3 - iter 84/146 - loss 0.41836824 - time (sec): 53.55 - samples/sec: 459.77 - lr: 0.000132 - momentum: 0.000000 2023-10-10 22:05:07,820 epoch 3 - iter 98/146 - loss 0.44258611 - time (sec): 63.06 - samples/sec: 473.68 - lr: 0.000131 - momentum: 0.000000 2023-10-10 22:05:17,349 epoch 3 - iter 112/146 - loss 0.42988452 - time (sec): 72.59 - samples/sec: 478.20 - lr: 0.000129 - momentum: 0.000000 2023-10-10 22:05:26,127 epoch 3 - iter 126/146 - loss 0.41794490 - time (sec): 81.36 - samples/sec: 480.97 - lr: 0.000127 - momentum: 0.000000 2023-10-10 22:05:34,136 epoch 3 - iter 140/146 - loss 0.41620303 - time (sec): 89.37 - samples/sec: 476.97 - lr: 0.000125 - momentum: 0.000000 2023-10-10 22:05:37,768 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:05:37,768 EPOCH 3 done: loss 0.4112 - lr: 0.000125 2023-10-10 22:05:43,626 DEV : loss 0.28306612372398376 - f1-score (micro avg) 0.1609 2023-10-10 22:05:43,635 saving best model 2023-10-10 22:05:44,623 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:05:54,251 epoch 4 - iter 14/146 - loss 0.29881808 - time (sec): 9.62 - samples/sec: 492.36 - lr: 0.000123 - momentum: 0.000000 2023-10-10 22:06:02,849 epoch 4 - iter 28/146 - loss 0.27924207 - time (sec): 18.22 - samples/sec: 471.98 - lr: 0.000121 - momentum: 0.000000 2023-10-10 22:06:12,815 epoch 4 - iter 42/146 - loss 0.34183632 - time (sec): 28.19 - samples/sec: 478.45 - lr: 0.000120 - momentum: 0.000000 2023-10-10 22:06:21,685 epoch 4 - iter 56/146 - loss 0.34079281 - time (sec): 37.06 - samples/sec: 471.65 - lr: 0.000118 - momentum: 0.000000 2023-10-10 22:06:30,830 epoch 4 - iter 70/146 - loss 0.33661860 - time (sec): 46.20 - samples/sec: 478.46 - lr: 0.000116 - momentum: 0.000000 2023-10-10 22:06:39,823 epoch 4 - iter 84/146 - loss 0.33761968 - time (sec): 55.20 - samples/sec: 477.34 - lr: 0.000115 - momentum: 0.000000 2023-10-10 22:06:50,180 epoch 4 - iter 98/146 - loss 0.32990070 - time (sec): 65.55 - samples/sec: 472.88 - lr: 0.000113 - momentum: 0.000000 2023-10-10 22:06:59,582 epoch 4 - iter 112/146 - loss 0.32645728 - time (sec): 74.96 - samples/sec: 464.23 - lr: 0.000111 - momentum: 0.000000 2023-10-10 22:07:08,979 epoch 4 - iter 126/146 - loss 0.32140043 - time (sec): 84.35 - samples/sec: 460.47 - lr: 0.000109 - momentum: 0.000000 2023-10-10 22:07:17,392 epoch 4 - iter 140/146 - loss 0.32290848 - time (sec): 92.77 - samples/sec: 457.95 - lr: 0.000108 - momentum: 0.000000 2023-10-10 22:07:21,205 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:07:21,205 EPOCH 4 done: loss 0.3174 - lr: 0.000108 2023-10-10 22:07:26,848 DEV : loss 0.2454710602760315 - f1-score (micro avg) 0.3563 2023-10-10 22:07:26,857 saving best model 2023-10-10 22:07:35,506 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:07:44,550 epoch 5 - iter 14/146 - loss 0.27566779 - time (sec): 9.04 - samples/sec: 454.76 - lr: 0.000105 - momentum: 0.000000 2023-10-10 22:07:54,215 epoch 5 - iter 28/146 - loss 0.33438474 - time (sec): 18.70 - samples/sec: 484.48 - lr: 0.000104 - momentum: 0.000000 2023-10-10 22:08:02,953 epoch 5 - iter 42/146 - loss 0.32225237 - time (sec): 27.44 - samples/sec: 484.25 - lr: 0.000102 - momentum: 0.000000 2023-10-10 22:08:11,715 epoch 5 - iter 56/146 - loss 0.29084459 - time (sec): 36.20 - samples/sec: 478.55 - lr: 0.000100 - momentum: 0.000000 2023-10-10 22:08:21,054 epoch 5 - iter 70/146 - loss 0.27706780 - time (sec): 45.54 - samples/sec: 481.60 - lr: 0.000099 - momentum: 0.000000 2023-10-10 22:08:30,720 epoch 5 - iter 84/146 - loss 0.26855614 - time (sec): 55.21 - samples/sec: 469.63 - lr: 0.000097 - momentum: 0.000000 2023-10-10 22:08:40,750 epoch 5 - iter 98/146 - loss 0.26372697 - time (sec): 65.24 - samples/sec: 463.02 - lr: 0.000095 - momentum: 0.000000 2023-10-10 22:08:51,290 epoch 5 - iter 112/146 - loss 0.26061632 - time (sec): 75.78 - samples/sec: 458.22 - lr: 0.000093 - momentum: 0.000000 2023-10-10 22:09:00,154 epoch 5 - iter 126/146 - loss 0.25838387 - time (sec): 84.64 - samples/sec: 456.78 - lr: 0.000092 - momentum: 0.000000 2023-10-10 22:09:08,987 epoch 5 - iter 140/146 - loss 0.25502563 - time (sec): 93.48 - samples/sec: 456.41 - lr: 0.000090 - momentum: 0.000000 2023-10-10 22:09:12,868 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:09:12,869 EPOCH 5 done: loss 0.2561 - lr: 0.000090 2023-10-10 22:09:18,945 DEV : loss 0.20666354894638062 - f1-score (micro avg) 0.4602 2023-10-10 22:09:18,955 saving best model 2023-10-10 22:09:25,879 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:09:35,820 epoch 6 - iter 14/146 - loss 0.21070114 - time (sec): 9.94 - samples/sec: 504.06 - lr: 0.000088 - momentum: 0.000000 2023-10-10 22:09:44,414 epoch 6 - iter 28/146 - loss 0.22631495 - time (sec): 18.53 - samples/sec: 483.31 - lr: 0.000086 - momentum: 0.000000 2023-10-10 22:09:53,224 epoch 6 - iter 42/146 - loss 0.21485580 - time (sec): 27.34 - samples/sec: 480.19 - lr: 0.000084 - momentum: 0.000000 2023-10-10 22:10:02,417 epoch 6 - iter 56/146 - loss 0.20592699 - time (sec): 36.53 - samples/sec: 482.40 - lr: 0.000083 - momentum: 0.000000 2023-10-10 22:10:11,772 epoch 6 - iter 70/146 - loss 0.20663057 - time (sec): 45.89 - samples/sec: 485.72 - lr: 0.000081 - momentum: 0.000000 2023-10-10 22:10:20,029 epoch 6 - iter 84/146 - loss 0.20369501 - time (sec): 54.15 - samples/sec: 478.12 - lr: 0.000079 - momentum: 0.000000 2023-10-10 22:10:28,856 epoch 6 - iter 98/146 - loss 0.20747226 - time (sec): 62.97 - samples/sec: 474.76 - lr: 0.000077 - momentum: 0.000000 2023-10-10 22:10:37,595 epoch 6 - iter 112/146 - loss 0.20151474 - time (sec): 71.71 - samples/sec: 474.56 - lr: 0.000076 - momentum: 0.000000 2023-10-10 22:10:46,686 epoch 6 - iter 126/146 - loss 0.20676841 - time (sec): 80.80 - samples/sec: 478.66 - lr: 0.000074 - momentum: 0.000000 2023-10-10 22:10:54,676 epoch 6 - iter 140/146 - loss 0.20652055 - time (sec): 88.79 - samples/sec: 474.05 - lr: 0.000072 - momentum: 0.000000 2023-10-10 22:10:58,983 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:10:58,983 EPOCH 6 done: loss 0.2048 - lr: 0.000072 2023-10-10 22:11:05,038 DEV : loss 0.18372002243995667 - f1-score (micro avg) 0.5195 2023-10-10 22:11:05,048 saving best model 2023-10-10 22:11:14,179 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:11:23,472 epoch 7 - iter 14/146 - loss 0.16630834 - time (sec): 9.29 - samples/sec: 515.70 - lr: 0.000070 - momentum: 0.000000 2023-10-10 22:11:31,585 epoch 7 - iter 28/146 - loss 0.15910112 - time (sec): 17.40 - samples/sec: 494.00 - lr: 0.000068 - momentum: 0.000000 2023-10-10 22:11:39,899 epoch 7 - iter 42/146 - loss 0.18159621 - time (sec): 25.72 - samples/sec: 496.78 - lr: 0.000067 - momentum: 0.000000 2023-10-10 22:11:48,970 epoch 7 - iter 56/146 - loss 0.16818074 - time (sec): 34.79 - samples/sec: 507.13 - lr: 0.000065 - momentum: 0.000000 2023-10-10 22:11:57,416 epoch 7 - iter 70/146 - loss 0.16285809 - time (sec): 43.23 - samples/sec: 506.75 - lr: 0.000063 - momentum: 0.000000 2023-10-10 22:12:05,581 epoch 7 - iter 84/146 - loss 0.16942539 - time (sec): 51.40 - samples/sec: 493.58 - lr: 0.000061 - momentum: 0.000000 2023-10-10 22:12:14,887 epoch 7 - iter 98/146 - loss 0.16837468 - time (sec): 60.70 - samples/sec: 487.16 - lr: 0.000060 - momentum: 0.000000 2023-10-10 22:12:24,198 epoch 7 - iter 112/146 - loss 0.16411960 - time (sec): 70.01 - samples/sec: 491.76 - lr: 0.000058 - momentum: 0.000000 2023-10-10 22:12:33,243 epoch 7 - iter 126/146 - loss 0.16260578 - time (sec): 79.06 - samples/sec: 490.77 - lr: 0.000056 - momentum: 0.000000 2023-10-10 22:12:42,356 epoch 7 - iter 140/146 - loss 0.16658286 - time (sec): 88.17 - samples/sec: 480.73 - lr: 0.000055 - momentum: 0.000000 2023-10-10 22:12:46,633 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:12:46,633 EPOCH 7 done: loss 0.1690 - lr: 0.000055 2023-10-10 22:12:52,591 DEV : loss 0.17367814481258392 - f1-score (micro avg) 0.604 2023-10-10 22:12:52,600 saving best model 2023-10-10 22:13:00,680 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:13:10,905 epoch 8 - iter 14/146 - loss 0.14405019 - time (sec): 10.21 - samples/sec: 481.65 - lr: 0.000052 - momentum: 0.000000 2023-10-10 22:13:20,300 epoch 8 - iter 28/146 - loss 0.15973513 - time (sec): 19.61 - samples/sec: 494.79 - lr: 0.000051 - momentum: 0.000000 2023-10-10 22:13:29,217 epoch 8 - iter 42/146 - loss 0.16052807 - time (sec): 28.53 - samples/sec: 486.59 - lr: 0.000049 - momentum: 0.000000 2023-10-10 22:13:38,038 epoch 8 - iter 56/146 - loss 0.15210947 - time (sec): 37.35 - samples/sec: 471.97 - lr: 0.000047 - momentum: 0.000000 2023-10-10 22:13:47,341 epoch 8 - iter 70/146 - loss 0.15143169 - time (sec): 46.65 - samples/sec: 474.18 - lr: 0.000045 - momentum: 0.000000 2023-10-10 22:13:56,442 epoch 8 - iter 84/146 - loss 0.15707453 - time (sec): 55.75 - samples/sec: 466.59 - lr: 0.000044 - momentum: 0.000000 2023-10-10 22:14:05,085 epoch 8 - iter 98/146 - loss 0.15161172 - time (sec): 64.39 - samples/sec: 465.47 - lr: 0.000042 - momentum: 0.000000 2023-10-10 22:14:14,292 epoch 8 - iter 112/146 - loss 0.14901872 - time (sec): 73.60 - samples/sec: 469.22 - lr: 0.000040 - momentum: 0.000000 2023-10-10 22:14:23,650 epoch 8 - iter 126/146 - loss 0.14657533 - time (sec): 82.96 - samples/sec: 464.50 - lr: 0.000039 - momentum: 0.000000 2023-10-10 22:14:32,792 epoch 8 - iter 140/146 - loss 0.14114562 - time (sec): 92.10 - samples/sec: 465.03 - lr: 0.000037 - momentum: 0.000000 2023-10-10 22:14:36,412 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:14:36,412 EPOCH 8 done: loss 0.1447 - lr: 0.000037 2023-10-10 22:14:42,561 DEV : loss 0.16653960943222046 - f1-score (micro avg) 0.6522 2023-10-10 22:14:42,570 saving best model 2023-10-10 22:14:52,366 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:15:01,070 epoch 9 - iter 14/146 - loss 0.10564763 - time (sec): 8.70 - samples/sec: 483.88 - lr: 0.000035 - momentum: 0.000000 2023-10-10 22:15:10,860 epoch 9 - iter 28/146 - loss 0.13546293 - time (sec): 18.49 - samples/sec: 494.59 - lr: 0.000033 - momentum: 0.000000 2023-10-10 22:15:19,875 epoch 9 - iter 42/146 - loss 0.13424335 - time (sec): 27.50 - samples/sec: 482.35 - lr: 0.000031 - momentum: 0.000000 2023-10-10 22:15:28,801 epoch 9 - iter 56/146 - loss 0.12517220 - time (sec): 36.43 - samples/sec: 481.07 - lr: 0.000029 - momentum: 0.000000 2023-10-10 22:15:37,873 epoch 9 - iter 70/146 - loss 0.12523915 - time (sec): 45.50 - samples/sec: 480.35 - lr: 0.000028 - momentum: 0.000000 2023-10-10 22:15:46,336 epoch 9 - iter 84/146 - loss 0.12571857 - time (sec): 53.97 - samples/sec: 479.53 - lr: 0.000026 - momentum: 0.000000 2023-10-10 22:15:55,269 epoch 9 - iter 98/146 - loss 0.13040790 - time (sec): 62.90 - samples/sec: 480.15 - lr: 0.000024 - momentum: 0.000000 2023-10-10 22:16:03,942 epoch 9 - iter 112/146 - loss 0.12849922 - time (sec): 71.57 - samples/sec: 479.88 - lr: 0.000023 - momentum: 0.000000 2023-10-10 22:16:13,165 epoch 9 - iter 126/146 - loss 0.12666524 - time (sec): 80.79 - samples/sec: 484.20 - lr: 0.000021 - momentum: 0.000000 2023-10-10 22:16:21,714 epoch 9 - iter 140/146 - loss 0.12857980 - time (sec): 89.34 - samples/sec: 479.99 - lr: 0.000019 - momentum: 0.000000 2023-10-10 22:16:25,263 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:16:25,264 EPOCH 9 done: loss 0.1282 - lr: 0.000019 2023-10-10 22:16:31,347 DEV : loss 0.1624951958656311 - f1-score (micro avg) 0.6638 2023-10-10 22:16:31,356 saving best model 2023-10-10 22:16:39,276 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:16:48,522 epoch 10 - iter 14/146 - loss 0.11168896 - time (sec): 9.24 - samples/sec: 485.28 - lr: 0.000017 - momentum: 0.000000 2023-10-10 22:16:57,095 epoch 10 - iter 28/146 - loss 0.12877886 - time (sec): 17.81 - samples/sec: 457.50 - lr: 0.000015 - momentum: 0.000000 2023-10-10 22:17:05,914 epoch 10 - iter 42/146 - loss 0.11898550 - time (sec): 26.63 - samples/sec: 463.39 - lr: 0.000013 - momentum: 0.000000 2023-10-10 22:17:15,949 epoch 10 - iter 56/146 - loss 0.10834056 - time (sec): 36.67 - samples/sec: 479.49 - lr: 0.000012 - momentum: 0.000000 2023-10-10 22:17:25,541 epoch 10 - iter 70/146 - loss 0.10628809 - time (sec): 46.26 - samples/sec: 483.48 - lr: 0.000010 - momentum: 0.000000 2023-10-10 22:17:34,310 epoch 10 - iter 84/146 - loss 0.10553800 - time (sec): 55.03 - samples/sec: 477.65 - lr: 0.000008 - momentum: 0.000000 2023-10-10 22:17:43,365 epoch 10 - iter 98/146 - loss 0.10631791 - time (sec): 64.08 - samples/sec: 474.69 - lr: 0.000007 - momentum: 0.000000 2023-10-10 22:17:52,344 epoch 10 - iter 112/146 - loss 0.10970536 - time (sec): 73.06 - samples/sec: 475.00 - lr: 0.000005 - momentum: 0.000000 2023-10-10 22:18:01,109 epoch 10 - iter 126/146 - loss 0.11345018 - time (sec): 81.83 - samples/sec: 473.11 - lr: 0.000003 - momentum: 0.000000 2023-10-10 22:18:10,282 epoch 10 - iter 140/146 - loss 0.11718872 - time (sec): 91.00 - samples/sec: 472.88 - lr: 0.000002 - momentum: 0.000000 2023-10-10 22:18:13,713 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:18:13,714 EPOCH 10 done: loss 0.1203 - lr: 0.000002 2023-10-10 22:18:19,724 DEV : loss 0.16049127280712128 - f1-score (micro avg) 0.6566 2023-10-10 22:18:20,641 ---------------------------------------------------------------------------------------------------- 2023-10-10 22:18:20,643 Loading model from best epoch ... 2023-10-10 22:18:24,871 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd 2023-10-10 22:18:37,888 Results: - F-score (micro) 0.6941 - F-score (macro) 0.5894 - Accuracy 0.5673 By class: precision recall f1-score support PER 0.7945 0.7443 0.7685 348 LOC 0.5994 0.7854 0.6799 261 ORG 0.3191 0.2885 0.3030 52 HumanProd 0.9091 0.4545 0.6061 22 micro avg 0.6736 0.7160 0.6941 683 macro avg 0.6555 0.5682 0.5894 683 weighted avg 0.6874 0.7160 0.6940 683 2023-10-10 22:18:37,888 ----------------------------------------------------------------------------------------------------